Intelligent kaolin gravel screening control method

Abstract

The invention relates to the field of gravel screening, in particular to an intelligent kaolin gravel screening control method, which is characterized in that a magnetic response energy difference function Ct is calculated, and a magnetic response threshold value Cth is combined for primary judgment, so that whether current particles are iron miscellaneous particles or not can be identified in real time, and then a track disturbance control mechanism is triggered. And calculating a trajectory intervention function Gy based on parameters such as a trajectory bending radius change rate, a magnetic field position function B (x, y), a magnetic disturbance direction included angle Jpi and the like, and performing secondary judgment on the trajectory intervention function Gy and a trajectory offset threshold value Gth to realize accurate control of effective offset of the particle trajectory. And the identified magnetic particle track is guided by utilizing the non-uniform magnetic disturbance field, so that layered discharge of the magnetic particle track and the main material flow is realized, and the separation success rate of the magnetic miscellaneous particles is remarkably improved. Compared with a traditional passive adsorption mode under a uniform magnetic field, the method can achieve active control and magnetic behavior intervention type screening path guiding with high predictability.

Description

Technical Field

[0001] The present invention relates to the technical field of sand and gravel screening, in particular to a kaolin sand and gravel intelligent screening control method. Background Art

[0002] The main raw materials of ceramics include clay, quartz and feldspar. Clay is one of the main raw materials of ceramics, mainly composed of hydrated aluminum silicate minerals such as kaolinite and montmorillonite, and its chemical composition includes aluminum oxide and silicon dioxide; the main component of quartz is silicon dioxide, which plays a skeleton role in ceramics, increasing the strength, hardness and heat resistance of ceramics; feldspar is a type of aluminum silicate mineral containing elements such as potassium, sodium, and calcium. It plays the role of a flux in the ceramic firing process, lowering the firing temperature of ceramics, promoting the melting of raw materials such as clay and quartz, forming a glass phase, and improving the density and mechanical properties of ceramics; in addition, ceramic raw materials also include kaolin, porcelain stone, colorants, blue and white materials, lime glaze and lime-alkali glaze, etc.

[0003] At present, the processing of high-performance ceramic materials belongs to the intersection of advanced materials engineering and intelligent powder manufacturing. In this field, the high-purity screening of kaolin sand and gravel is one of the key process links that determines the performance of ceramic products. Especially when manufacturing high-porcelain white ceramic firing materials, the control of the iron particle content in the sand and gravel is extremely strict, because iron impurities can easily produce oxidation and blackening reactions during high-temperature firing, seriously affecting the whiteness and light transmittance of the ceramic. At this stage, in the pretreatment process of sand and gravel raw materials in the existing ceramic industry, the removal of iron particles generally relies on traditional means such as uniform magnetic field adsorption, static vibration screening or manual gravity sorting. Such solutions have significant limitations: First, the magnetic field has a single mode of action, and only directional removal is achieved through magnetic adsorption, and the particle trajectory evolution path cannot be controlled. Second, the ability to identify weakly magnetic iron particles is insufficient, and it is easy to miss detection. Third, the screening behavior does not have a real-time feedback control mechanism, and it is difficult to adapt to the slight changes in the physical properties of the raw material particles. These technical bottlenecks directly lead to the mixing of iron impurities into the finished ceramic powder, further producing gray and black spots during the high-temperature firing process, destroying the visual performance and structural consistency of the "high-porcelain white" ceramic.

[0004] The above defects are mainly due to the lack of an integrated mechanism of "behavioral intervention and response judgment" in the screening process: iron particles do not always appear as dominant magnetic particles, and their microstructure may show a weak magnetic phase and unstable hysteresis response, resulting in their trajectory not changing significantly under a uniform magnetic field and being misjudged as "non-magnetic particles". In addition, the current magnetic screening device cannot distinguish between the root causes of "failure of trajectory disturbance" and "abnormal identification parameters", and the screening strategy cannot be adjusted, evaluated, or learned, which ultimately results in a large number of low-intensity magnetic particles not being removed during the screening process. This will directly lead to two types of abnormal consequences: first, the impurity iron content in kaolin exceeds the standard, resulting in local blackening after sintering, affecting the appearance and pass rate of ceramics; second, the equipment has a high misscreening rate, serious waste of resources, and cannot support the long-term production stability of high-precision ceramic manufacturing.

[0005] Therefore, how to intelligently screen kaolin sand and gravel is a technical problem that current technicians need to solve. Summary of the Invention

[0006] In view of the deficiencies in the prior art, the present invention provides a kaolin sand and gravel intelligent screening control method, which solves the problems mentioned in the background technology.

[0007] To achieve the above objectives, the present invention is implemented through the following technical solutions: a kaolin sand and gravel intelligent screening control method, comprising the following steps:

[0008] S1. Set sampling points on the screening drum and set sensor groups inside the sampling points to collect material screening data of kaolin sand and gravel particles in real time, and build a screening control system to wirelessly transmit the material screening data to the screening control system;

[0009] S2. In the screening control system, feature extraction is performed on the material screening data to obtain a de-ironization feature set, and the de-ironization feature set is pre-processed to obtain a standardized digital set;

[0010] S3. Based on the standardized digital set, calculate and output the magnetic response energy difference function Ct, and pre-set the magnetic response threshold Cth and the magnetic response energy difference function Ct for preliminary comparative evaluation;

[0011] S4. Based on the preliminary comparative evaluation, it is determined that the current kaolin sand and gravel particles are iron impurities. Then, trajectory perturbation and behavior guidance are triggered, and the trajectory intervention function Gy is calculated and output. The trajectory threshold Gth is preset and the trajectory intervention function Gy is compared and evaluated twice to classify and eliminate the trajectories in layers.

[0012] S5. Based on the magnetic response energy difference function Ct and the trajectory intervention function Gy, the effective screening index Esys is calculated and output, and the screening effect threshold Eth and the effective screening index Esys are set to comprehensively evaluate and analyze the effectiveness of de-ironization screening.

[0013] Preferably, said S1 includes S11 and S12;

[0014] S11. Setting sampling points in the screening drum and setting sensor groups in the sampling points to collect material screening data of kaolin sand and gravel particles in real time during the screening process;

[0015] The sampling points are set at the feed port, magnetic disturbance field area, conventional particle area, track deviation area and discharge port of the screening drum;

[0016] The sensor group includes a micro high-frequency magnetic induction coil array, a signal conditioner, a dual-channel magnetic sensitive probe detection, a high-speed CCD track monitoring sensor, a Tof laser imaging sensor and a MEMS magnetic field probe array;

[0017] The material screening data includes voltage data, magnetic response peak time, magnetic flux change start time, particle displacement data, particle size, reflectivity, particle image and magnetic induction intensity value;

[0018] S12. Integrate the sensor group with the screening drum central control module and build a screening control system at the same time. Set up LaRo wireless communication technology through the wireless communication module of the screening drum central control module, wirelessly connect the wireless communication module with the screening control system, and transmit the real-time material screening data to the screening control system.

[0019] Preferably, said S2 includes S21 and S22;

[0020] S21. receiving material screening data in real time in the screening control system, and performing feature extraction on the material screening data to obtain a de-ironization feature set;

[0021] The de-ironization feature set includes the perturbation spin reversal frequency Vs, the local magnetic response delay Em, the impurity recognition potential factor Zf, the trajectory bending radius k of the i-th particle in the r-th motion trajectory ri , the magnetic induction intensity value B(x i ,y i ) and the relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi ;

[0022] The perturbation spin reversal frequency Vs is obtained by extracting the voltage data variation period through spectrum analysis of the voltage data, and converted into the response frequency of the reversal of the micro-magnetic domain inside the particle per unit time;

[0023] The local magnetic response delay Em is extracted by performing difference calculation between the peak time of the magnetic response and the start time of the magnetic flux change;

[0024] The impurity identification potential factor Zf is obtained by training a machine learning model based on a data set of identified iron impurities, taking the perturbation spin reversal frequency Vs, local magnetic response delay Em, particle size, reflectivity and particle image as features, inputting them into the machine learning model, and outputting the impurity identification potential factor Zf of potential iron impurities;

[0025] The trajectory bending radius k of the i-th particle in the r-th trajectory ri By positioning the continuous trajectory of kaolin sand and gravel particles at the pixel level according to the particle displacement data, the discrete curvature formula is used to fit the curve segment and then deduce and extract;

[0026] The magnetic induction intensity value B(x i ,y i ) By analyzing the magnetic induction intensity value of each kaolin sand and gravel particle, a magnetic field map is formed through a DEM simulation framework to extract the magnetic induction intensity value B, and the horizontal axis of the pixel point of the magnetic field map is used as the horizontal axis coordinate x, and the vertical axis of the pixel point is used as the vertical axis coordinate y;

[0027] The relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi Extracted by analyzing the angle between the magnetic disturbance direction and the particle trajectory direction;

[0028] S22. Based on the obtained de-ironization feature set, pre-processing is performed to obtain a standardized digital set, wherein the pre-processing includes a Z-Score standardization process to eliminate the unit dimension effect of all parameters in the de-ironization feature set.

[0029] Preferably, said S3 includes S31 and S32;

[0030] S31, start the screening drum to screen the kaolin sand and gravel particles, and start the sensor group and magnetic disturbance to extract the perturbation spin reversal frequency Vs, local magnetic response delay Em, impurity identification potential factor Zf and the relative vibration reversal angle J of the i-th particle in the p-th magnetic disturbance direction in the standardized digital set. pi , the magnetic response energy difference function Ct is calculated and output in the screening control system to analyze the magnetic response behavior of kaolin sand and gravel particles at different angles.

[0031] Preferably, S32, by calculating and outputting a magnetic response energy difference function Ct for a batch of iron particles and non-magnetic particles, and performing an average calculation, outputting a magnetic response threshold value Cth, and performing a preliminary comparative evaluation of the magnetic response energy difference function Ct obtained in real time and the magnetic response threshold value Cth to determine the current iron particle situation in the kaolin sand and gravel particles, the specific evaluation content is as follows;

[0032] When the magnetic response energy difference function Ct ≥ the magnetic response threshold Cth, it means that the current i-th particle has magnetic response behavior and is preliminarily judged to be an iron particle. At this time, trajectory perturbation and behavior guidance are triggered;

[0033] When the magnetic response energy difference function Ct is less than the magnetic response threshold Cth, it means that the current i-th particle is a non-magnetic particle. At this time, the pulsed air flow nozzle of the screening drum is triggered instantaneously to blow the kaolin sand and gravel particles into the conventional particle area and discharge them through the discharge port of the conventional particle area, retaining the iron particles preliminarily judged in the magnetic disturbance field area.

[0034] Preferably, said S4 includes S41 and S42;

[0035] S41. When the current kaolin sand and gravel particles are initially judged to be iron impurities by preliminary comparison and evaluation, the trajectory perturbation and behavior guidance are triggered. The trajectory perturbation and behavior guidance are performed by extracting the trajectory bending radius k of the i-th particle in the r-th motion trajectory from the standardized digital set. ri , the magnetic induction intensity value B(x i ,y i ) and the relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi , calculate and output the trajectory intervention function Gy.

[0036] Preferably, S42, for a batch of particles that have successfully and unsuccessfully suppressed magnetic perturbation deviation, sample the trajectory intervention function Gy, analyze the statistical differences, and take the 75% quantile of the differences as the trajectory threshold Gth. A secondary comparative evaluation is performed between the trajectory intervention function Gy obtained in real time and the trajectory threshold Gth to determine the particle trajectory deviation under magnetic perturbation. The specific evaluation content is as follows;

[0037] When the trajectory intervention function Gy ≥ the trajectory threshold Gth, it means that the magnetic disturbance is effective. At this time, the current particle is guided to the de-ironization channel by the magnetic disturbance;

[0038] When the trajectory intervention function Gy is less than the trajectory threshold Gth, it indicates that the magnetic disturbance is invalid, and an abnormality is preliminarily determined, and the current particle is returned to the normal particle area.

[0039] Preferably, said S5 includes S51, S52 and S53;

[0040] S51. Comprehensively calculate the effective magnetic disturbance trajectory intervention function Gy and the magnetic response energy difference function Ct of the initially determined iron particles to output the effective screening index Esys, and analyze the actual iron removal response.

[0041] Preferably, S52, based on the user's high-porcelain white ceramic firing standard, a screening effect threshold Eth is set, and then the screening effect threshold Eth is comprehensively evaluated with the effective screening index Esys obtained in real time to analyze the effectiveness of the current de-ironing screening. The specific evaluation content is as follows;

[0042] When the effective screening index Esys ≥ screening effect threshold Eth, it means that the iron removal effect meets the user-set standard, and the screening is successful.

[0043] When the effective screening index Esys is less than the screening effect threshold Eth, it means that the iron removal effect does not meet the user-set standard, and the optimization control strategy is executed at this time.

[0044] Preferably, S53, the optimization control strategy includes screen drum speed adjustment and magnetic disturbance optimization;

[0045] The screen drum speed is adjusted by adopting a three-stage speed rotation, dynamically adjusting the speed by ±10% every 10 minutes. The current speed causes the particles to be subjected to excessive centrifugal force, and the magnetic disturbance force is insufficient to guide their deviation trajectory. The high-speed rotation of the particles causes the trajectory to be stable, which weakens the magnetic disturbance intervention effect.

[0046] The magnetic disturbance optimization improves the magnetic source in the magnetic disturbance field area by 40%.

[0047] The present invention provides a method for intelligent screening and control of kaolin sand and gravel, which has the following beneficial effects:

[0048] (1) This method sets sampling points on the screening drum and combines it with a sensor group to build a material screening data collection system with multi-point sampling throughout the entire process, effectively acquiring key screening data. Relying on the screening control system, the above data is feature extracted to construct a de-ironization feature set, which is then standardized through Z-score to form a standardized digital set. This provides an accurate and stable input basis for subsequent magnetic response judgment and behavioral intervention, thereby improving the accuracy of iron particle identification and the robustness of data processing from the source.

[0049] (2) This method calculates the magnetic response energy difference function Ct and combines it with the magnetic response threshold Cth for initial judgment, which can instantly identify whether the current particle is an iron impurity particle, thereby triggering the trajectory disturbance control mechanism. The trajectory intervention function Gy is calculated based on parameters such as the trajectory bending radius change rate, the magnetic field position function B(x,y), and the magnetic disturbance direction angle Jpi, and a secondary judgment is made with the trajectory offset threshold Gth to achieve precise control of whether the particle trajectory is effectively offset. The identified magnetic particle trajectory is guided by a non-uniform magnetic disturbance field to achieve layered discharge from the main material flow, thereby significantly improving the success rate of magnetic impurity sorting. Compared with the traditional passive adsorption method under a uniform magnetic field, this method can achieve active control and highly predictable magnetic behavior intervention screening path guidance.

[0050] (3) Based on the identification of particle magnetism and behavioral deviation, this method proposes a fusion judgment mechanism - the effective screening index Esys, and combines it with the screening effect threshold Eth to achieve a comprehensive evaluation of the system's de-ironization effect. When the system's operating status does not meet the de-ironization standard set by the user, the feedback control strategy will be automatically triggered. By adjusting the three-stage screen drum speed and increasing the magnetic source intensity in the magnetic disturbance field area, two core parameters, real-time self-optimization adjustment can be completed during operation without changing the equipment structure. This feedback mechanism effectively overcomes external interference such as raw material magnetic fluctuations and particle property variations, ensuring the continuous stability and evaluation traceability of the system's de-ironization ability, and adapting to the stringent requirements of high-end materials such as high-porcelain white ceramics for de-ironization process accuracy. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a schematic diagram of the steps of a kaolin sand and gravel intelligent screening control method of the present invention;

[0052] Figure 2 Schematic diagram comparing the trajectory deviations of iron particles and non-magnetic particles in the magnetic disturbance field area. DETAILED DESCRIPTION

[0053] The following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the accompanying drawings. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts are within the scope of protection of the present invention.

[0054] Example 1

[0055] The present invention provides a kaolin sand and gravel intelligent screening control method, please refer to Figure 1 and Figure 2 , including the following steps:

[0056] S1. Set sampling points on the screening drum and set sensor groups inside the sampling points to collect material screening data of kaolin sand and gravel particles in real time, and build a screening control system to wirelessly transmit the material screening data to the screening control system;

[0057] S2. In the screening control system, feature extraction is performed on the material screening data to obtain a de-ironization feature set, and the de-ironization feature set is pre-processed to obtain a standardized digital set;

[0058] S3. Based on the standardized digital set, calculate and output the magnetic response energy difference function Ct, and pre-set the magnetic response threshold Cth and the magnetic response energy difference function Ct for preliminary comparative evaluation;

[0059] S4. Based on the preliminary comparative evaluation, it is determined that the current kaolin sand and gravel particles are iron impurities. Then, trajectory perturbation and behavior guidance are triggered, and the trajectory intervention function Gy is calculated and output. The trajectory threshold Gth is preset and the trajectory intervention function Gy is compared and evaluated twice to classify and eliminate the trajectories in layers.

[0060] S5. Based on the magnetic response energy difference function Ct and the trajectory intervention function Gy, the effective screening index Esys is calculated and output, and the screening effect threshold Eth and the effective screening index Esys are set to comprehensively evaluate and analyze the effectiveness of de-ironization screening.

[0061] In the present embodiment, the method adopts a screening control process with high-frequency magnetic response perception, trajectory behavior modeling, dual threshold discrimination and feedback optimization as the core. By arranging sampling points in multiple areas within the screening drum and combining a sensor group, the material screening data of kaolin sand and gravel particles in the screening process are collected in real time. The data completes the extraction of multidimensional features in the screening control system and forms a digital feature set after standardization. Then, a magnetic response energy difference function Ct is constructed based on this, and a preliminary comparison is made with the set magnetic response threshold Cth to determine whether the particles have ferromagnetic response behavior. For particles initially judged to be iron particles, their trajectory intervention function Gy is further calculated and compared with the trajectory offset threshold Gth. If some pre-qualified particles are effective, they are guided into the magnetic discharge channel. Otherwise, backflow screening is performed to achieve accurate identification and path diversion of magnetic and non-magnetic particle trajectory behaviors. Finally, the effective screening index Esys is output by the magnetic response energy difference function Ct and the trajectory intervention function Gy, and compared with the screening standard Eth set by the user, to achieve system evaluation and control feedback of the overall de-ironization effect. Through this method, a closed-loop control mechanism from particle identification, behavior guidance and effect evaluation is realized. Compared with traditional means based on physical adsorption or static screening, this method significantly improves the recognition ability of weakly magnetic iron particles, trajectory stratification accuracy and de-ironification success rate, and is particularly suitable for the high purity requirements in the screening process of kaolin raw materials for high-porcelain white ceramics. The system also has dynamic control capabilities, and can achieve adaptive optimization of the de-ironization effect by adjusting the magnetic disturbance intensity and the screen drum speed, thereby ensuring the robustness and stability of the screening system under raw material fluctuation conditions. Therefore, the present invention has significant technical advantages in improving the de-ironization purity, reducing the misscreening rate, and enhancing the level of system automation.

[0062] Example 2

[0063] See also Figure 1 ,Specifically: S1 includes S11 and S12;

[0064] S11. Setting sampling points in the screening drum and setting sensor groups in the sampling points to collect material screening data of kaolin sand and gravel particles in real time during the screening process;

[0065] Sampling points are set at the feed port, magnetic disturbance field area, conventional particle area, trajectory deviation area and discharge port of the screening drum. Among them, the magnetic disturbance field area is used for magnetic deironing. The trajectory deviation area represents the spatial area where the trajectory of kaolin sand and gravel particles undergoes quantifiable deviation from the original path due to the superposition of magnetic disturbance guidance force, centrifugal force, gravity and collision force after passing through the non-uniform magnetic field area. It is set after the magnetic disturbance field area and before the discharge port. It is the core reaction space for trajectory twisting, stratification and diversion of kaolin screening materials.

[0066] The sensor group includes a miniature high-frequency magnetic induction coil array, a signal conditioner, a dual-channel magnetic sensitive probe detection, a high-speed CCD track monitoring sensor, a Tof laser imaging sensor and a MEMS magnetic field probe array;

[0067] Material screening data includes voltage data, magnetic response peak time, magnetic flux change start time, particle displacement data, particle size, reflectivity, particle image and magnetic induction intensity value;

[0068] Among them, voltage data is collected by using a micro high-frequency magnetic induction coil array arranged in the magnetic disturbance field area. When kaolin sand and gravel particles pass through the magnetic disturbance field area, the iron particles in the kaolin sand and gravel particles will disturb the parameter small voltage change, which is collected through the signal conditioner;

[0069] The magnetic response peak time and the magnetic flux change start time are measured by setting a dual-channel magnetic sensitive probe at the edge of the magnetic disturbance field area, where channel 1 measures the magnetic flux change start time to obtain the magnetic flux change start time, and channel 2 monitors its magnetic response peak time;

[0070] Particle displacement data, particle size, reflectivity and particle images are acquired by combining a CCD track monitoring sensor with a Tof laser imaging sensor in the track deviation area.

[0071] The magnetic induction intensity value is obtained by the MEMS magnetic field probe array in the track deviation area;

[0072] S12. Integrate the sensor group with the screening drum central control module and build a screening control system at the same time. Set up LaRo wireless communication technology through the wireless communication module of the screening drum central control module, wirelessly connect the wireless communication module with the screening control system, and transmit the real-time material screening data to the screening control system.

[0073] In this embodiment, the method sets multiple functional sampling points inside the screening drum, covering key screening nodes such as the feed port, magnetic disturbance field area, conventional particle area, trajectory deviation area and discharge port, and deploys a multi-type sensor group consisting of a micro high-frequency magnetic induction coil array, a dual-channel magnetic sensitive probe, a high-speed CCD trajectory monitoring sensor, a ToF laser imaging sensor and a MEMS magnetic field probe at each sampling point. The sensor group can collect multimodal material screening data including voltage response signals, magnetic flux change time points, particle displacement, particle size, reflectivity, image information and magnetic induction intensity in real time, with high sensitivity, high timeliness and multi-dimensional data fusion capabilities. In particular, the micro-inductor device in the magnetic disturbance field area monitors the instantaneous voltage changes caused by iron particles, and the dual-channel magnetic sensitive probe captures the dynamic process of magnetic response at two time points, thereby achieving early detection of the response behavior of weakly magnetic particles. Subsequently, the above-mentioned collected raw material screening data is wirelessly integrated into the screening control system through the central control module. The system achieves stable, low-latency wireless data transmission by integrating LaRo communication technology, ensuring data integrity and real-time performance under high-frequency sampling conditions, and providing a structured and continuous input data stream for subsequent feature extraction and intelligent screening decisions.

[0074] Example 3

[0075] See also Figure 1 , specifically: S2 includes S21 and S22;

[0076] S21. receiving material screening data in real time in the screening control system, and performing feature extraction on the material screening data to obtain a de-ironization feature set;

[0077] The de-ironization feature set includes the perturbation spin reversal frequency Vs, the local magnetic response delay Em, the impurity recognition potential factor Zf, the trajectory bending radius k of the i-th particle in the r-th motion trajectory, and the local magnetic response delay Em. ri , the magnetic induction intensity value B(x i ,y i ) and the relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi ;

[0078] The perturbation spin reversal frequency Vs is calculated by extracting the voltage data variation period through spectral analysis of the voltage data and converting it into the response frequency of reversal of the micro-magnetic domains within the particle per unit time. When iron impurity particles are disturbed by a non-uniform magnetic field, the internal magnetic domains will produce spin reversal or reverse perturbation phenomena. The higher this frequency, the more active the particle may be in magnetic behavior, which is significantly different from the non-magnetic particles in ceramic kaolin materials.

[0079] The local magnetic response delay Em is extracted by performing difference calculation between the peak time of the magnetic response and the start time of the magnetic flux change. After the particle enters the magnetic disturbance field, its magnetic response will be delayed. Due to the complex internal magnetic structure of iron particles, the response to the magnetic field will lag. The delayed response is an important manifestation of highly magnetic particles and can assist in identifying weak iron particles.

[0080] The impurity identification potential factor Zf is obtained by training a machine learning model based on a data set of identified iron impurities. The perturbation spin reversal frequency Vs, local magnetic response delay Em, particle size, reflectivity, and particle image are input into the machine learning model to output the impurity identification potential factor Zf of potential iron impurities.

[0081] The trajectory bending radius k of the i-th particle in the r-th trajectory ri By positioning the continuous trajectory of kaolin sand and gravel particles at the pixel level according to the particle displacement data, the discrete curvature formula is used to fit the curve segment and then deduce and extract;

[0082] The magnetic induction intensity value B(x i ,y i ) By analyzing the magnetic induction intensity value of each kaolin sand and gravel particle, a magnetic field map is formed through a DEM simulation framework to extract the magnetic induction intensity value B, and the horizontal axis of the pixel point of the magnetic field map is used as the horizontal axis coordinate x, and the vertical axis of the pixel point is used as the vertical axis coordinate y;

[0083] The relative vibration reversal angle J of the i-th particle in the p-th magnetic disturbance direction pi By analyzing the angle between the magnetic disturbance direction and the particle trajectory, the angle is extracted and used to adjust the magnetic field direction layout to ensure that the main magnetic disturbance direction coincides with the target trajectory to the greatest extent possible.

[0084] S22. Based on the obtained de-ironization feature set, pre-processing is performed to obtain a standardized digital set. The pre-processing includes a strange normalization process, and the unit dimension effect of all parameters in the de-ironization feature set is eliminated by using a Z-Score normalization method.

[0085] In this embodiment, the method receives the original screening data of kaolin sand and gravel particles from multiple sampling points through a screening control system, and constructs a de-ironification feature set based on the behavioral response of the particles in the magnetic disturbance field. The feature set integrates six types of highly sensitive indicators. The six-dimensional de-ironification feature set not only comprehensively covers the triple attributes of response, behavior and position of magnetic particles, but also integrates the empirical induction of historical data samples by machine learning, so that the recognition algorithm has good generalization ability and data adaptability. In order to eliminate the influence of inconsistent units, unequal dimensions and scale differences among the above-mentioned multi-dimensional features on the accuracy of subsequent discrimination calculations, the Z-Score standardization method is used to normalize and pre-process all feature parameters, and output a standardized digital set. This standardization process ensures that the subsequent magnetic response function and trajectory intervention function calculation models have rigorous mathematical consistency and comparability, providing a solid data foundation for achieving high-precision identification evaluation and hierarchical screening.

[0086] Example 4

[0087] See also Figure 1 ,Specifically: S3 includes S31 and S32;

[0088] S31, start the screening drum to screen the kaolin sand and gravel particles, and start the sensor group and magnetic disturbance to extract the perturbation spin reversal frequency Vs, local magnetic response delay Em, impurity identification potential factor Zf and the relative vibration reversal angle J of the i-th particle in the p-th magnetic disturbance direction in the standardized digital set. pi , calculate and output the magnetic response energy difference function Ct in the screening control system, and analyze the magnetic response behavior of kaolin sand and gravel particles at different angles;

[0089] The magnetic response energy difference function Ct is calculated and output by the following algorithm formula;

[0090] Ct i =Em i ·Vs i ·Zf i ·cos(J pi );

[0091] Where Ct i represents the magnetic response energy difference function of the i-th particle, Em i represents the local magnetic response delay of the i-th particle, Vs i represents the perturbation spin reversal frequency of the i-th particle, Zf i represents the impurity recognition potential factor of the i-th particle, and cos represents the cosine function;

[0092] The significance of the formula is that if the kaolin sand and gravel particles have a large response delay under magnetic disturbance, absorb a lot of magnetic energy, have a high spin frequency, an active micromagnetic structure, a high probability of identifying potential iron impurities, and a stable offset direction and clear behavior, it means that the kaolin sand and gravel particles may be iron impurities.

[0093] S32. Calculate and output a magnetic response energy difference function Ct for a sample of iron particles and non-magnetic particles, perform an average calculation, and output a magnetic response threshold value Cth. A preliminary comparative evaluation is performed between the magnetic response energy difference function Ct obtained in real time and the magnetic response threshold value Cth to determine the current iron particle situation in the kaolin sand and gravel particles. The specific evaluation content is as follows;

[0094] When the magnetic response energy difference function Ct ≥ the magnetic response threshold Cth, it means that the current i-th particle has magnetic response behavior and is preliminarily judged to be an iron particle. At this time, trajectory perturbation and behavior guidance are triggered;

[0095] When the magnetic response energy difference function Ct is less than the magnetic response threshold Cth, it means that the current i-th particle is a non-magnetic particle. At this time, the pulsed air flow nozzle of the screening drum is triggered instantaneously to blow the kaolin sand and gravel particles into the conventional particle area and discharge them through the discharge port of the conventional particle area, retaining the iron particles preliminarily judged in the magnetic disturbance field area.

[0096] In this embodiment, the method dynamically screens kaolin sand and gravel particles by starting the screening drum, and simultaneously activates the sensor group and its data stream processing module to extract multi-dimensional key parameters from the standardized digital set in real time. The screening control system calculates and outputs the magnetic response energy difference function Ct based on the above parameters, where each parameter corresponds to the particle's response activity to magnetic disturbance, hysteresis characteristics, impurity possibility and magnetic disturbance direction coupling degree in a physical sense. The magnetic response energy difference function Ct is quantitatively evaluated using a structured multi-factor coupling model, which can comprehensively analyze the response characteristics of a single particle under a multi-angle magnetic disturbance field to determine whether it has typical iron impurity characteristics. At the same time, the historical magnetic response energy difference function Ct is calculated through a large number of known iron impurity particles and non-magnetic particle samples, and the mean statistics are performed to extract the magnetic response threshold Cth as the initial reference standard for current particle identification. During the screening process, the magnetic response energy difference function Ct value of each particle obtained in real time will be compared with the magnetic response threshold Cth: when the magnetic response energy difference function Ct ≥ the magnetic response threshold Cth, it means that the particle shows significant magnetic response characteristics in the magnetic disturbance field and is preliminarily identified as an iron particle. The system will trigger the subsequent trajectory perturbation and behavior guidance steps; if the magnetic response energy difference function Ct < the magnetic response threshold Cth, it is considered to be a non-magnetic particle, and the pulsed airflow nozzle built into the screening drum is triggered at the same time to quickly guide the particle to the regular particle area and discharge it through the discharge port to avoid accidentally entering the magnetic disturbance screening channel. Through this embodiment, the present invention establishes a high-precision intelligent screening model with magnetic response as the main factor, feature quantization as the basis, and threshold comparison as the decision-making mechanism in the initial stage of de-ironization identification. It not only achieves effective identification of weakly magnetic and fine-grained iron particles, but also has real-time response capabilities and stable discrimination logic, significantly improving the screening efficiency and accuracy of magnetic impurities, reducing misscreening and missed judgments, and has a significant effect on improving the screening purity and processing efficiency of kaolin raw materials for high-porcelain white ceramics.

[0097] Example 5

[0098] See also Figure 1 and Figure 2 , specifically: S4 includes S41 and S42;

[0099] S41. When the current kaolin sand and gravel particles are initially judged to be iron impurities by preliminary comparison and evaluation, the trajectory perturbation and behavior guidance are triggered. The trajectory perturbation and behavior guidance are performed by extracting the trajectory bending radius k of the i-th particle in the r-th motion trajectory from the standardized digital set. ri , the magnetic induction intensity value B(x i ,y i ) and the relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi , calculate and output trajectory intervention function Gy;

[0100] The trajectory intervention function Gy is calculated and outputted by the following algorithm formula;

[0101]

[0102] Where Gy i represents the trajectory intervention function of the i-th particle, d represents the integral variable, dt represents the time integral variable, u i represents the magnetic susceptibility of the i-th particle. When the kaolin sandstone particle passes through the magnetic disturbance field area, the micro-inductance probe reads the magnetic induction intensity disturbance value caused by the particle entering the magnetic disturbance field area. The magnetic induction intensity disturbance value is calculated by ratioing the theoretical magnetic field intensity at that point in space and then dimensionlessly processed.

[0103] The physical meaning of the formula is that when a particle enters the magnetic disturbance zone, if the particle is a non-magnetic material, such as hydromica and quartz, there will be no obvious interference with the magnetic field; if the particle is a magnetic material, such as iron impurities, it will enhance or disturb the local magnetic flux density. In the formula, if the particle trajectory suddenly bends significantly or the magnetic field force is strong, the particle magnetic response is low, and the direction of magnetic disturbance is opposite to the direction of particle movement, it means that the particle has been successfully interfered with by the magnetic disturbance and guided into the magnetic de-ironization channel in the magnetic disturbance field area;

[0104] Indicates the rate of change of the trajectory bending radius. A larger value indicates that the trajectory is undergoing rapid bending and deviation.

[0105] It represents the magnetic effect term. The stronger the magnetic field and the weaker the magnetic particles, the stronger the guiding ability.

[0106] cos(J pi ) The closer the angle between the direction of magnetic disturbance and the direction of particle motion is to 0°, that is, the larger the cosine is, the more effective the interference is.

[0107] S42. For a batch of particles that have successfully and unsuccessfully suppressed magnetic perturbation deviation, sample the trajectory intervention function Gy, analyze the statistical differences, and take the 75% quantile of the differences as the trajectory threshold Gth. Then, perform a secondary comparative evaluation of the trajectory intervention function Gy obtained in real time and the trajectory threshold Gth to determine the particle trajectory deviation under magnetic perturbation. The specific evaluation content is as follows;

[0108] When the trajectory intervention function Gy ≥ the trajectory threshold Gth, it means that the magnetic disturbance is effective. At this time, the current particle is guided to the de-ironization channel by the magnetic disturbance;

[0109] When the trajectory intervention function Gy is less than the trajectory threshold Gth, it indicates that the magnetic disturbance is invalid, and an abnormality is preliminarily determined, and the current particle is returned to the normal particle area.

[0110] In this embodiment, when a particle is judged as an iron particle in the preliminary evaluation of the magnetic response energy difference function Ct, the method immediately enters the trajectory perturbation and behavior guidance. The screening control system extracts the trajectory bending radius k of the particle i in the rth motion trajectory from the standardized digital set. ri , the magnetic induction intensity value B(x i ,y i ) and the relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi Calculate and output the trajectory intervention function Gy, and comprehensively determine whether the particle has experienced a quantifiable trajectory deviation in the magnetic perturbation field. This function physically characterizes whether the magnetic perturbation force has effectively interfered with the particle's motion path, reflecting the combined effect of the particle's trajectory change rate, the directionality of the magnetic field's guiding force, and the strength of the magnetic perturbation behavior. Based on the distribution of the trajectory intervention function Gy values for "successful" and "unsuccessful" particles in historical screening samples, the difference is statistically analyzed and the 75th percentile is extracted as the trajectory deviation threshold Gth, which serves as the baseline for determining trajectory deviation behavior. During the screening process, if the real-time calculated trajectory intervention function Gy ≥ the trajectory deviation threshold Gth, the magnetic perturbation intervention behavior is effective, and the system guides the particle into the magnetic de-ironization channel for removal. If the trajectory intervention function Gy is less than the trajectory deviation threshold Gth, the intervention fails to guide the particle, and the intervention is deemed invalid, returning the particle to the regular particle area for rescreening. This implementation not only determines whether the particle exhibits magnetic response during the identification phase, but also performs secondary verification of its trajectory changes during the behavioral intervention phase, thus achieving a dual closed-loop control mechanism of "identification and behavior." The beneficial effects of this mechanism are reflected in: clearly separating the misjudgment situations of "strong magnetic response" but "no deviation", preventing particles with unstable trajectories from mistakenly entering the de-ironing channel; improving the objectivity and accuracy of trajectory diversion judgment, and enhancing the control certainty of stratified discharge behavior; reducing the de-ironing deviation caused by ineffective disturbance, significantly improving the magnetic impurity particle discharge efficiency and system intelligence level, and providing key behavioral layer support for high-precision magnetic screening.

[0111] Example 6

[0112] See also Figure 1 , specifically: S5 includes S51, S52 and S53;

[0113] S51, comprehensively calculating the effective magnetic disturbance trajectory intervention function Gy and the magnetic response energy difference function Ct of the initially determined iron particles, outputting the effective screening index Esys, and analyzing the actual iron removal response;

[0114] The effective screening index Esys is calculated and output by the following algorithm formula;

[0115]

[0116] Where N represents the total number of kaolin sand and gravel particles, δ(·) represents the logic function used to determine whether the conditions are met, Gy avg represents the average value of the trajectory intervention function Gy, Ct avg Represents the average value of the magnetic response energy difference function Ct.

[0117] S52. Based on the user's high-porcelain white ceramic firing standard, a screening effect threshold Eth is set, and then the screening effect threshold Eth is comprehensively evaluated with the effective screening index Esys obtained in real time to analyze the effectiveness of the current de-ironing screening. The specific evaluation content is as follows;

[0118] When the effective screening index Esys ≥ screening effect threshold Eth, it means that the iron removal effect meets the user-set standard, and the screening is successful.

[0119] When the effective screening index Esys is less than the screening effect threshold Eth, it means that the iron removal effect does not meet the user-set standard, and the optimization control strategy is executed at this time.

[0120] S53, optimization control strategy including screen drum speed adjustment and magnetic disturbance optimization;

[0121] The screen drum speed is adjusted by adopting a three-stage speed rotation, dynamically adjusting the speed by ±10% every 10 minutes. The current speed causes the particles to be subjected to excessive centrifugal force, and the magnetic disturbance force is insufficient to guide their deviation trajectory. The high-speed rotation of the particles causes the trajectory to be stable, which weakens the magnetic disturbance intervention effect.

[0122] Magnetic perturbation optimization increases the magnetic source in the magnetic perturbation field area by 40%, resulting in detectable deviations in more particle trajectories.

[0123] In this embodiment, the method fuses the magnetic response energy difference function Ct of the magnetic response iron particles identified in the previous stage with the trajectory intervention function Gy corresponding to the actual trajectory deviation behavior, and calculates the output effective screening index Esys in a statistical weighted manner. This index not only takes into account the magnetic potential Ctavg of the particles, but also integrates its deviation performance Gyavg in the magnetic disturbance field. Through the logic function, the particles that meet the magnetic identification and effective trajectory deviation during the screening process are extracted, and normalized and quantitatively analyzed, and finally a measurable system screening quality index is formed. Based on the user's personalized requirements for the iron removal rate in the high-porcelain white ceramic firing process, the screening effect threshold Eth is preset and compared with the Esys calculated in real time. When the effective screening index Esys ≥ the screening effect threshold Eth, it means that the current iron removal screening effect has met or exceeded the set standard, and a screening success prompt is output; when the effective screening index Esys < the screening effect threshold Eth, it indicates that the current magnetic particle removal effect does not meet the standard, and the optimization control strategy is automatically triggered to enter the parameter self-adjustment process. The optimized control strategy uses a dual-parameter control strategy to ensure that the system can quickly restore iron removal performance under non-ideal conditions: First, by implementing a three-stage dynamic rotation control of the screen drum speed, the speed is adjusted by ±10% every 10 minutes, effectively adjusting the balance between centrifugal force and magnetic disturbance force, breaking the stable inertia of particle trajectories, and improving the ability to intervene in magnetic disturbances. Second, by increasing the magnetic source power in the magnetic disturbance field area by approximately 40%, the spatial gradient and intensity of the magnetic field are enhanced, causing more particle trajectories to undergo quantifiable deviations, thereby improving the removal efficiency of magnetically responsive particles. Through this implementation, a closed-loop control chain is achieved, from "particle behavior identification" to "screening result feedback" to adaptive control. The beneficial effects of this mechanism are reflected in: the effective screening index Esys, as a comprehensive performance evaluation indicator, improves the accuracy of screening effect judgment and the targetedness of system adjustment; the introduction of the user-set threshold Eth makes it configurable, flexible and process adaptable; the optimized control strategy realizes real-time adjustment based on real feedback data, significantly improving the robustness, stability and accuracy maintenance ability of screening under complex material fluctuation conditions, providing stable and reliable intelligent screening technology support for the high-purity iron removal needs of the ceramic industry.

[0124] While embodiments of the present invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations may be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A kaolin sand and gravel intelligent screening control method, characterized in that: The following steps are involved: S1. Setting sampling points on the screening drum and setting sensor groups inside the sampling points to collect material screening data of kaolin sand and gravel particles in real time. The material screening data includes voltage data, magnetic response peak time, magnetic flux change start time, particle displacement data, particle size, reflectivity, particle image and magnetic induction intensity value; S2. Extract features from the material screening data to obtain a de-ironization feature set, and pre-process the de-ironization feature set to obtain a standardized digital set; S3. Based on the standardized digital set, a magnetic response energy difference function Ct is calculated, and the magnetic response energy difference function Ct is preliminarily compared and evaluated with a preset magnetic response threshold value Cth, and a preliminary comparison and evaluation result is output; S4. Based on the preliminary comparative evaluation results, if the current kaolin sand and gravel particles are determined to be iron impurities, trajectory perturbation and behavior guidance are triggered, and a trajectory intervention function Gy is calculated and output. The trajectory intervention function Gy is then subjected to a secondary comparative evaluation with a preset trajectory threshold Gth to determine trajectory classification and stratified elimination. S5. Based on the magnetic response energy difference function Ct and the trajectory intervention function Gy, the effective screening index Esys is calculated and output, and the effective screening index Esys and the screening effect threshold Eth are comprehensively evaluated and analyzed to finally determine the effectiveness of kaolin sand and gravel deironing screening.

2. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: In said S1, said sampling points are set at the feed port, magnetic disturbance field area, conventional particle area, track deviation area and discharge port of the screening drum; The sensor group includes a micro high-frequency magnetic induction coil array, a signal conditioner, a dual-channel magnetic sensitive probe detection, a high-speed CCD track monitoring sensor, a Tof laser imaging sensor and a MEMS magnetic field probe array.

3. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: In S2, the de-ironization feature set includes the perturbation spin reversal frequency Vs, the local magnetic response delay Em, the impurity recognition potential factor Zf, the trajectory bending radius k of the i-th particle in the r-th motion trajectory, and the local magnetic response delay Em. ri , the magnetic induction intensity value B(x i ,y i ) and the relative vibration reverse angle J of the i-th particle in the p-th magnetic disturbance direction pi ; Among them, the perturbation spin reversal frequency Vs is the voltage data change period extracted by spectrum analysis of voltage data, which is converted into the response frequency of reversal of micromagnetic domains inside the particles per unit time; the impurity recognition potential factor Zf is obtained by training a machine learning model based on the data set of identified iron impurities; the local magnetic response delay Em is obtained by extracting the difference between the peak time of the magnetic response and the start time of the magnetic flux change; the trajectory bending radius k ri The pixel-level positioning of the continuous trajectory of kaolin sandstone particles is determined by particle displacement data; the magnetic induction intensity value B (x i ,y i ) is determined based on the magnetic induction intensity value of each kaolin sand and gravel particle; the relative vibration reverse angle J pi It is determined by analyzing the angle between the direction of magnetic disturbance and the direction of particle trajectory.

4. The intelligent screening control method for kaolin sand and gravel according to claim 3, characterized in that: In S3, the magnetic response energy difference function Ct is calculated using the following algorithm formula: Ct i =Em i ·Vs i ·Zf i ·cos(J pi ); Where Ct i represents the magnetic response energy difference function of the i-th particle, Em i represents the local magnetic response delay of the i-th particle, Vs i represents the perturbation spin reversal frequency of the i-th particle, Zf i represents the impurity recognition potential factor of the i-th particle, and cos represents the cosine function.

5. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: In S3, a preliminary comparison and evaluation is performed on the magnetic response energy difference function Ct and a preset magnetic response threshold Cth, including: When the magnetic response energy difference function Ct ≥ the magnetic response threshold Cth, it means that the current i-th particle has magnetic response behavior and is preliminarily judged to be an iron particle, and the trajectory perturbation and behavior guidance are triggered; When the magnetic response energy difference function Ct is less than the magnetic response threshold Cth, it means that the current i-th particle is a non-magnetic particle, and at the same time, the pulsed airflow nozzle of the screening drum is instantaneously triggered to blow the kaolin sand and gravel particles to the conventional particle area, and retain the iron particles preliminarily judged in the magnetic disturbance field area.

6. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: In S4, the calculation formula of the trajectory intervention function Gy is: Where Gy i represents the trajectory intervention function of the i-th particle, d represents the integral variable, dt represents the time integral variable, u i represents the magnetic susceptibility of the i-th particle; Indicates the rate of change of trajectory bending radius; represents the magnetic force term. The stronger the magnetic field and the weaker the particle magnetism, the stronger the guiding ability. cos(J pi ) represents the angle between the magnetic disturbance direction and the particle motion direction.

7. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: In S4, the trajectory intervention function Gy is subjected to a secondary comparative evaluation with a preset trajectory threshold Gth, including: When the trajectory intervention function Gy ≥ the trajectory threshold Gth, it means that the magnetic disturbance is effective, and the current particle is guided to the de-ironization channel by the magnetic disturbance; When the trajectory intervention function Gy is less than the trajectory threshold Gth, it indicates that the magnetic disturbance is invalid, the preliminary judgment is determined to be abnormal, and the current particle is returned to the normal particle area.

8. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: The effective screening index Esys is calculated by the following algorithm formula: Where N is the total number of kaolin sand particles, δ(·) is a logical function used to determine whether the conditions are met, Gy avg represents the average value of the trajectory intervention function Gy, Ct avg Represents the average value of the magnetic response energy difference function Ct.

9. The intelligent screening control method for kaolin sand and gravel according to claim 1, characterized in that: In S5, the effective screening index Esys and the screening effect threshold Eth are comprehensively evaluated and analyzed, including: When the effective screening index Esys ≥ screening effect threshold Eth, it means that the iron removal effect meets the user-set standard, and the screening is successful. When the effective screening index Esys is less than the screening effect threshold Eth, it means that the iron removal effect does not meet the user-set standard, and the optimization control strategy is executed at this time.

10. The intelligent screening control method for kaolin sand and gravel according to claim 9, characterized in that: The optimization control strategy includes screen drum speed adjustment and magnetic disturbance optimization; The screen drum speed is adjusted by adopting a three-stage speed rotation, dynamically adjusting the speed by ±10% every preset time. When the current speed causes the particles to be subjected to excessive centrifugal force and the magnetic disturbance force is insufficient to guide their deviation trajectory, it is determined that the high-speed rotation of the particles stabilizes the trajectory and weakens the magnetic disturbance intervention effect. The magnetic disturbance optimization is achieved by improving the magnetic source in the magnetic disturbance field area.

Images